The investigators propose to determine eukaryotic initiation factor 4E, which assists RNA helicases to bind to 51 untranslated regions of mRNA. Overexpression in cell lines produces a malignant phenotype. In preliminary studies, overexpression determined by Western blot occurs in breast cancers, but not in benign breast lesions. Preliminary data also associates the degree of overexpression with tumor aggressiveness and survival. However, these preliminary studies are based on small numbers. Thus, the investigators propose to perform a prospective study of 168 women with stage II or III breast cancer. Their first aim is to assess eIF4E overexpression in breast tissue, and their second aim is to perform ER and PR analysis so as to be able to relate this data to the eIF4E. They believe that in 5 years, they will have enough statistical power to determine if overexpression is associated with recurrence. The investigators believe that this data might lead to more informed decisions regarding who needs chemotherapy, i.e., by implication that if eIF4E is predictive of recurrence in stage II, then it might be for stage I as well. Then, a much larger study with sufficient power for determining recurrence in stage I breast cancer will be justified.